An Overview of Data Mining Techniques Applied for Heart ... · An Overview of Data Mining...

An Overview of Data Mining Techniques

Applied for Heart Disease Diagnosis and

Prediction

Salha M. Alzahani, Afnan Althopity, Ashwag Alghamdi, Boushra Alshehri, and Suheer Aljuaid Dept. of Computer Science, College of Computers and Information Technology

Taif University, Taif, Saudi Arabia

Email: [email protected]

Abstract — Data mining techniques have been applied

magnificently in many fields including business, science, the

Web, cheminformatics, bioinformatics, and on different

types of data such as textual, visual, spatial, real-time and

sensor data. Medical data is still information rich but

knowledge poor. There is a lack of effective analysis tools to

discover the hidden relationships and trends in medical data

obtained from clinical records. This paper reviews the state-

of-the-art research on heart disease diagnosis and

prediction. Specifically in this paper, we present an

overview of the current research being carried out using the

data mining techniques to enhance heart disease diagnosis

and prediction including decision trees, Naive Bayes

classifiers, K-nearest neighbour classification (KNN),

support vector machine (SVM), and artificial neural

networks techniques. Results show that SVM and neural

networks perform positively high to predict the presence of

coronary heart diseases (CHD). Decision trees after features

reduction is the best recommended classifier to diagnose

cardiovascular disease (CVD). Still the performance of data

mining techniques to detect coronary arteries diseases (CAD)

is not encouraging (between 60%-75%) and further

improvements should be pursued.

Index Terms—heart disease, data mining, decision tree,

naive bayes, K-nearest neighbor, support vector machine

I. INTRODUCTION

Knowledge discovery in data is defined as: “the

extraction of hidden previously unknown and potentially

useful information about data" [1]. Basically knowledge

discovery in data is the process of extracting different

features from data in various steps. Fig.1 shows the

process of Knowledge discovery from various data

sources in a specific domain. Data mining is the heart

(core) step, which results in the discovery of implicit but

potentially valuable knowledge from huge amount of data.

Data mining technology provides the user with the

methods to find new and implicit patterns from massive

data. In the healthcare domain, discovered knowledge can

be used by the healthcare administrators and medical

physicians to improve the accuracy of diagnosis, to

enhance the goodness of surgical operations and to reduce

Manuscript received July 18, 2014; revised December 15, 2014.

the harmful effects of drug [2], [3]. It aims also to propose

less expensive therapeutic [4].

Figure 1. Process of knowledge discovery in data.

The diagnosis of diseases is a difficult but critical task

in medicine. The detection of heart disease from “various

factors or symptoms is a multi-layered issue which is not

free from false presumptions often accompanied by

unpredictable effects” [5]. Thus, we can use patients’ data

that have been collected and recorded to ease the

diagnosis process and utilize knowledge and experience of

numerous specialists dealt with the same symptoms of

diseases. Providing invaluable services with less costs is a

major constraint by the healthcare organizations (hospitals,

polyclinics, and medical centres). According to [6],

“valuable quality service denotes the accurate diagnosis of

patients and providing efficient treatment. Poor clinical

decisions may lead to disasters and hence are seldom

entertained”. Besides, it is essential that the hospitals

decrease the cost of clinical tests. Using professional and

expert computerized systems based on machine-learning

and data mining methods should help in one direction or

another with achieving clinical tests or diagnosis at

reduced risks [7], [8].

This paper aims to provide a survey of current

techniques of knowledge discovery using data mining

techniques applied to medical research; particularly, to

heart disease prediction. Literature studies between 2010

and 2014 are discussed, unless a significant study before

that should be mentioned. A number of experiments and

research works have been done to compare the

performance of predictive data mining techniques like

decision tree, Naive Bayes, K-nearest neighbour, support

vector machine and artificial neural networks. This paper

discussed the results of the state-of-the-art techniques and

gives conclusions towards future research.

310©2014 Engineering and Technology Publishing

Lecture Notes on Information Theory Vol. 2, No. 4, December 2014

doi: 10.12720/lnit.2.4.310-315

II. HEART DISEASES: AN OVERVIEW

The heart can be affected by diverse types of diseases

most of them are dangerous on human lives. Coronary

heart diseases, cardiomyopathy and cardiovascular

diseases are some examples of heart diseases, as shown in

Fig. 2.

Figure 2. Heart arteries miseases.

The most common type of these diseases is Coronary

Arteries Disease (CAD) wherein coronary arteries hard

and tight [9]. The term Cardiovascular Disease (CVD)

denotes a wide range of conditions that affect the heart

and the blood vessels, and the manner in which the blood

is pumped and circulated throughout the body [4], [10].

CVD results in severe illness, disability, and is most likely

to cause death. Narrowing of the coronary arteries causes

the reduction of oxygen supplied to the heart and leads to

the so-called Coronary Heart Disease (CHD) [4], [10]. A

sudden blockage of a coronary artery is generally due to a

blood clot which may cause a heart attack. Chest pains

arise when the blood received by the heart muscles is

inadequate and unconnected [4].

There are many remarks and symptoms used by the

physicians to diagnose heart diseases. Age, sex, chest pain

type, blood pressure, cholesterol, fasting blood sugar,

maximum heart rate, and hereditary are meaningful

symptoms [6], [11]. Besides, other habits might be used

including stress, overweight, smoking, alcohols intake and

less exercise [6].

III. DATA MINING TECHNIQUES IN HEALTH CARE

After exploring some types of heart arteries diseases

and symptoms that when have certain values denote a

heart disease, in this section we will explore different data

mining techniques applied generally to healthcare.

Symptoms and patient records can be used as features and

such huge amounts of data can be used for knowledge

discovery in the health care domain. A general framework

proposed by [2] for medical data mining is shown in Fig.

3. The framework starts with a specific medical problem

wherein a dataset should be pre-processed and cleaned

before mining the data using one of the available data

mining tools. Knowledge evaluation comes at the last and

expertise from the medical domain should involve.

Figure 3. Framework for medical data mining

A. Neural Networks

Neural networks are biologically inspired highly

interconnected cells that simulate the human brain [1].

The perceptron is the simplest architecture which has one

neuron and a learning method. More sophisticated

architecture is multi-layer neural networks (MLP) which

one or more neurons connected at different layers. Neural

networks can be trained to learn a classification task and

to predict diseases [6], [12].

B. K-nearest Neighbor Algorithm (KNN)

K-nearest neighbour classification algorithm is a well-

known method for classifying an unseen instance using

the classification of the instances closest to it [1]. Basic

KNN classification algorithm works by finding K training

instances that are close to the unseen instance using

distance measures such as Euclidean, Manhattan,

maximum dimension distance, and others. Then, the

algorithm decides the class for the unseen instance by

taking the most commonly occurring class in the nearest

K instances [1].

C. Decision Tree Classification Algorithm

Decision trees are powerful classification algorithms

used alternatively as decision/classification rules [1].

Popular decision tree algorithms include Quinlan’s ID3,

C4.5, C5, and Breiman et al.’s CART [13]-[15]. As its

name implies, this classification technique works by

recursively constructing branches of the tree based on

certain observations (or variables). A well-known

algorithm to construct the branches is called top-down

induction of decision trees (TDIDT) [1]. Decision trees

are easily constructed with binary or categorical variables

but the mission becomes harder with numerical variables.

A corresponding threshold value must be specified for the

later based on some mathematical or observational

considerations in order to be able to construct the

branches of the tree. This step is repeated at each leaf

node until the complete tree is constructed ending with

leaves which gives one of the classes or predictions in the

dataset. The objective of the splitting algorithm is “to find

a variable-threshold pair that maximizes the homogeneity

of the resulting two or more subgroups of samples” [16].

1) C4.5 classification algorithm

One of the best decision tree algorithms is C4.5. This

algorithm can manage continuous data in numerical forms

using pruning algorithms which aim to simplify the



classification rules without any loss of prediction accuracy.

Only the most important features are kept whereby they

lowered the error rates. One of its factors is denoted by M

which indicates the minimum instances that a leaf should

have. C means the confidence threshold which is

considered for pruning. By changing these two factors, the

accuracy of algorithm can be increased and the error can

be decreased [9].

2) RIPPER classification algorithm

RIPPER stands for Repeated Incremental Pruning to

Produce Error Reduction. This classification algorithm

which was proposed by Cohen [17], is based on

association rules with reduced error pruning (REP), a very

common and effective technique found in decision tree

algorithms [16]. To generate association rules using REP

algorithm, the training data is divided into a growing set

and a pruning set. The growing set is the initial

association rules which can be generated purely from the

dataset using some heuristic methods. The growing set

contains a huge set of rules that should be repeatedly

simplified to form the pruning set. Thus, the simplification

is done using typical pruning operators which may allow

to delete a term from any single rule or from different

association rules. The pruning operator chosen for

simplification should give the most accurate rule with the

greatest reduction of error. The simplification process

terminates when applying the pruning operator would

increase the error value on the pruning set [16].

D. Support Vector Machine (SVM)

SVM is a state-of-the-art maximum margin

classification algorithm rooted in statistical learning

theory [16]. It is a method for classification of both linear

and non-linear data. The training data is converted into n-

dimensional data using non-linear transformation method.

Then, the algorithm searches for the best hyper-plane to

separate the transformed data into two different classes.

SVM performs classification tasks by maximizing the

margin of the hyper-plane separating both classes while

minimizing the classification errors.

E. Naïve Bayes Algorithm

One of the Bayesian methods is Naïve Bayes classifiers

which uses the probabilistic formula:

)(

)()|()|(

BP

APABPBAP

where A and B are two events (e.g. the probability that the

train will arrive on time given that the weather is rainy).

Such Naïve Bayes classifiers use the probability theory to

find the most likely classification of an unseen

(unclassified) instance [1]. The algorithm performs

positively with categorical data but poorly if we have

numerical data in the training set [9].

IV. DATA MINING TECHNIQUES FOR HEART DISEASE

DIAGNOSIS AND PPREDICTION

In this section, a survey of medical data mining

techniques applied for diagnosis and prediction of some

types of heart diseases is presented. Literature studies

from 2010 and above are discussed, unless a significant

study before that should be mentioned. Symptoms

denoting a heart disease are processed. Standard datasets

for each heart disease are prepared including certain

features each with a predefined range of values. Such

datasets have been used for knowledge discovery of heart

problems in several research studies [4]-[11], [15], [16],

[18]-[20].

As mentioned in Section II, the most common type of

heart diseases is CAD (Coronary Arteries Disease). For

diagnosing CAD, different classification algorithms have

been implemented [9]-[20]. In these studies, three features

of vessel’s stenosis have been employed namely Left

Anterior Descending (LAD), Left Circumflex (LCX) and

Right Coronary Artery (RCA). The ones whom LAD,

LCX or RCA vessel is clogged are classified as CAD

patients, others as healthy. Table I summarises the results

obtained from different classification algorithms for

diagnosis of CAD specifically. In [20], neural networks algorithm was used for

prediction of stenosis of each vessel separately. Ten-fold cross-validation methods were involved to measure the accuracy. A multi-layered perceptron neural network was employed for the classification. The accuracy reached 73%, 64.85% and 69.39% for LAD, LCX and RCA vessels, respectively.

Another study [9] applied Naïve Bayes, C4.5, and

KNN classification algorithms using more sophisticated

features that have not been applied in [20]. The study

aimed to diagnose CAD via the stenosis of each LAD,

LCX or RCA vessel separately. RapidMiner tool was used

and the accuracy was measured using 10-fold cross

validation technique. The best accuracy was obtained by

C4.5 wherein achieved accuracy were 74.20%, 63.76%,

and 68.33%, respectively. The accuracy obtained using

C4.5 is the ideal one for diagnosing CAD via LAD

stenosis and was not achieved by previous studies.

Another types of heart attacks are CVD and CHD

explained briefly in Section II. Different studies have

explored the prediction of these heart diseases but unlike

previous studies, without considering the stenosis of each

LAD, LCX or RCA vessel separately. Table II shows the

results obtained from different studies to diagnose CVD

and CHD diseases. In [10], CVD and CHD were predicted

using three different supervised machine learning

algorithms namely Naïve Bayes, KNN, decision list

techniques. Tanagra tool was used to classify the data, the

experiments were conducted using 10-fold cross

validation, and the results were compared. The accuracy

from Naïve Bayes, KNN, and decision list were 52.33%,

52%, 45.67%, respectively. Naive Bayes algorithm

showed slightly better performance compared with the

other algorithms.

Moreover, CVD heart disease prediction were further

analysed using four data mining classification techniques

namely RIPPER classifier, decision tree, artificial neural

network, and SVM [16]. The results were compared using

10-fold cross validation method and the accuracy obtained

were as follows: 81.08%, 79.05%, 80.06%, and 84.12%,

respectively. Their analysis shows that out of these four

classification models, SVM predicts cardiovascular



disease with the highest accuracy. A study by [19] used

genetic algorithms (GA) to reduce the actual data size by

reducing the number of attributes used in previous studies.

Thus, it aims to get the optimal subset of attributed

sufficient for CVD prediction. The study then applied

three classifiers namely Naïve Bayes, decision tree, and

classification via clustering. As clustering is the process of

grouping similar instances into different groups or clusters,

it was used as a pre-processing step before feeding the

data to the classifying model. Experiments were

conducted using WEKA 3.6.0 tool. The results obtained

from the three classifiers outperform previous results in

the literature reaching 96.5%, 99.2%, and 88.3,

respectively. Another study [21] used Naïve Bayes and

decision trees to detect CVD and achieved 62.03%, and

60.40% of accuracy, respectively.

TABLE I. ACCURACY OF NEURAL NETWORK, NAÏVE BAYES, C4.5 AND

KNN CLASSIFICATION ALGORITHMS FOR DIAGNOSIS OF CORONARY

ARTERIES DISEASE (CAD)

Data Mining

Techniques

Accuracy

(LAD)

Accuracy

(LCX)

Accuracy

(RCA)

Ref.

Neural

Network 73% 64.85% 69.39%

Babaoglu et

al. [20]

Naïve Bayes 51.81% 62.73% 67.29% Alizadehsani

et al. [9] C4.5 74.20% 63.76% 68.33%

KNN 59.65% 61.39% 59.11%

a. Studies that consider LAD, LCX, and RCA vessel separately to

diagnose CAD

TABLE II. ACCURACY OF DECISION TREE, NAÏVE BAYES AND

CLASSIFICATION VIA CLUSTERING ALGORITHMS FOR DIAGNOSIS OF

CARDIOVASCULAR DISEASE (CVD) AND CORONARY HEART DISEASE

(CHD)

Data Mining Techniques Accuracy Ref.

Naïve Bayes 52.33% Rajkumar and

Reena [10] KNN 52%

Decision List 45.67%

RIPPER Classifier 81.08%

Kumari and

Godara [16]

Decision Tree 79.05%

Artificial Neural Network 80.06%

SVM 84.12%

Naïve Bayes 96.5%

Anbarsi et al. [19]

Decision Tree 99.2%

Classification via Clustring 88.3%

Naïve Bayes 62.03% Sitar-Taut, et al.

[21] Decision Tree 60.40%

Hybrid Genetic Neural Network 89% Amin et al. [22]

C4.5 82.5%

Srinivas et al. [23]

Multilayer Neural Network (MLP) 89.7%

Bayesian Classifier 82%

SVM 82.5%

C5 Classifier 89.6% AbuKhousa and

Campbell [12] MLP 91.0%

SVM 92.1%

a. Studies that do not consider LAD, LCX, and RCA vessel

separately to diagnose CVD and CHD diseases.

A recent study [22] used a hybrid genetic neural

network method to detect CVD heart disease and achieved

89% of accuracy.

On the other hand, a number of studies have explored

CHD heart diseases particularly. A study to predict CHD

were employed using WEKA and applying four data

mining techniques namely C4.5, multilayer neural

network (MLP), Bayesian classifier, and SVM [23]. The

results obtained showed that out of these four models,

MLP obtained the highest accuracy (89.7%) compared

with the others. Another three prediction models for CHD

were implemented in [12] namely C5 classifier, MLP and

SVM. The results obtained from 10-forld cross validation

method showed that SVM was the best predictor gaining

the accuracy of 92.1%.

TABLE III. EFFECTIVENESS OF DATA MINING TECHNIQUES USED FOR

HEART DISEASE DIAGNOSIS AND PREDICTION

Data Mining

Techniques Purpose of study

Maximum

Accuracy Ref.

Neural Network To diagnose the

presence of CHD. 91.0%

AbuKhous

a and Campbell

[12]

Naïve Bayes

Classifier

+GA Feature Reduction

To diagnose the

presence of CVD. 96.5%

Anbarsi at

el. [19]

Decision Tree

+GA Feature Reduction

To diagnose the


Anbarsi at

el. [19]

RIPPER

Classifier

To diagnose the


Kumari

and

Godara [16]

C4.5 Classifier

To diagnose CAD

via stenosis of LAD vessel

74.20% Alizadehsa

ni et al. [9]

To diagnose the


Srinivas et

al. [23]

C5 Classifier To diagnose the


AbuKhous

a and

Campbell [12]

SVM To diagnose the


AbuKhous

a and Campbell

[12]

Clustering To diagnose the


Anbarsi at el. [19]

KNN

To diagnose CAD

via stenosis of

LCX vessel.

61.39% Alizadehsani et al. [9]

Hybrid Genetic

Neural Network

To diagnose the

presence of CVD. 89%

Amin et al.

[22]

a. Maximum accuracy obtained by each technique to diagnose a certain type of heart disease.

V. DISCUSSION

As you have seen, the performance of different data

mining algorithms on the heart disease datasets has been

studied and compared. The accuracy of these techniques

have been measured using 10-fold cross validation

method on a standard dataset of each disease. Almost the

same features have been employed for the prediction of

each type of disease. Therefore, these techniques can be

analysed and compared on a larger scale as you will see

below.

In this section, several medical data mining techniques

for heart diseases diagnosis focusing on three famous

types namely CAD, CVD and CHD are discussed. The

effectiveness of different techniques are shown in Table



III. For each technique, the maximum accuracy rate

obtained for a certain disease is shown. For example, the

maximum accuracy rate obtained using neural network is

91.0% to predict CHD by AbuKhousa and Campbell [12]

whereas Naïve Bayes classifier achieved the maximum

accuracy of 96.5% to predict CVD by Anbarsi at el. [19].

The table shows ten different techniques and their

maximum accuracy rate for heart disease prediction (i.e.

patient classification as healthy or affected by certain

disease).

Based on such evaluation, the highest and uppermost

accuracy to diagnose each type of disease using a certain

data mining technique can be investigated (shown in bold

font in the table). We can recommend the following: (i)

C4.5 classifier perform better than other data mining

techniques to diagnose CAD via stenosis of the LAD

vessel, followed by KNN via the stenosis of LCX vessel.

(ii) SVM and neural networks perform comparably and

positively high; therefore, they can be utilised to predict

the presence of CHD. (iii) Decision trees method after the

reduction and optimization of features using GA is the

best recommended classifier to diagnose CVD heart

disease.

Figure 4. Comparison of different medical data mining techniques for heart diseases prediction.

Fig. 4 shows the performance evaluation metrics of

data mining techniques for heart disease prediction. It is

observed that the accuracy of various classification

techniques for CVD diagnosis is highly encouraging

(between 85% and 99%). Consequently, diagnosis

systems that employs classifiers or clusters can assist the

medical professionals in making decision about CVD

early diagnosis. Moreover, data mining techniques

perform positively well with diagnosing CHD (achieving

accuracy between 82% and 92%). Still the performance of

classification methods to detect CAD diseases is not

encouraging (between 60%-75%) whether or not the LAD,

LCX, and RCA vessels are considered separately.

Therefore, further research should be carried out using

more sophisticated features and hybrid algorithms to

improve the prediction of CAD diseases.

Due to the successful implementation of data mining

techniques for heart diagnosis, various heart disease

prediction systems that employed the aforementioned data

mining techniques such as Intelligent Heart Disease

Prediction System (IHDPS) [7] have been proposed.

However, there some limitations in these systems. One of

these weaknesses is the fact that physicians do not feel

comfortable or having a good knowledge of using

computerized systems for diagnosis. A recent survey [24]

showed that existing systems are computational expensive

and do not achieve optimum accuracy nor 100% reliable

results.

VI. CONCLUSION AND FUTURE WORK

Early diagnosis of heart diseases may save humans

from heart attacks. This paper reviews the state-of-the-art

data mining techniques applied for diagnosing three heart

diseases namely CAD, CVD, and CHD. Among the

famous and sever diseases is CAD which can be

diagnosed via the stenosis of blood vessels. Such disease

is data rich but unfortunately the obtained accuracy from

CAD classifiers is poor. Data mining techniques applied

for CVD and CHD are promising. Results showed that the

optimization and feature reduction utilising GA or

principle component analysis (PCA) for a certain disease

may strongly increase the accuracy of a classifier. It is

found that decision trees and Naïve Bayes classifiers are

recommended for CVD diagnosis with an accuracy

reaching more than 95%. Further, C5, SVM, and neural

networks are the best recommended classifiers for CHD

prediction. It is observed that the prediction results of

various data mining classification techniques are strongly

encouraging and would assist the physicians to do early

diagnosis and make more accurate decisions. Nonetheless,

data mining techniques do not achieve 100% accuracy for

heart diseases prediction and hence cannot be utilised

solely for diagnosis.

Future works should focus on improving the

predication of CAD diseases utilising more features and

separate combinations of vessel stenosis. Furthermore,

feature reduction should be utilised in various ways to

achieve better accuracy results with all diseases. New

classifiers should be developed for other heart diseases

and problems such as coronary microvascular diseases,

pulmonary, and cyanotic heart diseases. This work can be

further extended by working with different heart related

datasets from health care organizations and agencies using

all the available techniques and also using a combination

of them.

REFERENCES

[1] M. Bramer, Principles of Data Mining: Springer-Verlag, 2007.

[2] Z. Jitao and W. Ting, "A general framework for medical data

mining," presented at the International Conference on Future Information Technology and Management Engineering (FITME),

2010.

[3] S. Tsumoto, "Problems with mining medical data," presented at the The 24th Annual International Computer Software and

Applications Conference, 2000.

[4] K. Srinivas, B. K. Rani, and A. Govardhan, "Applications of data mining techniques in healthcare and prediction of heart attacks,"

International Journal on Computer Science and Engineering, vol. 2, pp. 250-255, 2010.

[5] S. B. Patil and Y. S. Kumaraswamy, "Extraction of significant

patterns from heart disease warehouses for heart attack prediction," International Journal of Computer Science and

Networks Security, vol. 9, pp. 228-235, 2009.

[6] S. Oyyathevan and A. Askarunisa, "An expert system for heart disease prediction using data mining technique: Neural network,"



International Journal of Engineering Research and Sports Science, vol. 1, pp. 1-6, 2014.

[7] S. Palaniappan and R. Awang, "Intelligent heart disease prediction

system using data mining techniques," presented at the International Conference on Computer Systems and Applications,

2008.

[8] G. Subbalakshmi, "Decision support in heart disease prediction system using naive bayes," Indian Journal of Computer Science

and Engineering, vol. 2, pp. 170-174, 2011.

[9] R. Alizadehsani, J. Habibi, B. Bahadorian, H. Mashayekhi, A. Ghandeharioun, and R. Boghrati, et al., "Diagnosis of coronary

arteries stenosis using data mining," J Med Signals Sens, vol. 2, pp.

153-9, Jul 2012. [10] A. Rajkumar and G. S. Reena, "Diagonsis of heaer disease using

data mining algorithm," Global Journal of Computer Science and

Technology, vol. 10, pp. 38-43, 2010. [11] M. A. Jabbar, B. L. Deekshatulu, and P. Chandra, "Graph based

approach for heart disease prediction," in Proc. Third

International Conference on Trends in Information, Telecommunication and Computing, vol. 150, 2013, pp. 465-474.

[12] Y. W. Xing, J. Wang, Z. H. Zhao, and Y. H. Gao., "Combination

data mining methods with new medical data to predicting outcome of coronary heart disease," presented at the International

Conference on Convergence Information Technology, 2007.

[13] L. Rokach and O. Maimon, "Top-down induction of decision trees classifiers - a survey," IEEE Transactions on Systems, Man, and

Cybernetics, Part C: Applications and Reviews, , vol. 35, pp. 476-

487, 2005. [14] S. R. Safavian and D. Landgrebe, "A survey of decision tree

classifier methodology," IEEE Transactions on Systems, Man and

Cybernetics, vol. 21, pp. 660-674, 1991. [15] S. Ranganatha, H. R. P. Raj, C. Anusha, and S. K. Vinay,

"Medical data mining and analysis for heart disease dataset using

classification techniques," presented at the National Conference on Challenges in Research & Technology in the Coming Decades,

2013.

[16] M. Kumari and S. Godara, "Comparative study of data mining classification methodsin cardiovascular disease prediction," IJCST

vol. 2, pp. 304-305, 2011. [17] W. W. Cohen, "Fast effective rule induction," presented at the

Twelfth International Conference on Machine Learning, 1995.

[18] R. Alizadehsani, J. Habibi, Z. Alizadeh Sani, H. Mashayekhi, R. Boghrati, and A. Ghandeharioun, et al., "Diagnosing coronary

artery disease via data mining algorithms by considering

laboratory and echocardiography features," Res Cardiovasc Med, vol. 2, pp. 133-139, 2013.

[19] M. Anbarsi, E. Anupriya, and N. Iyengar, "Enhanced prediction of heart disease with feature subset selection using genetic

algorithm," International Journal of Engineering Science and

Technology, vol. 2, pp. 5370-5376, 2010. [20] I. Babaoglu, O. K. Baykan, N. Aygul, K. Ozdemir, and M. Bayrak,

"Assessment of exercise stress testing with artificial neural

network in determining coronary artery disease and predicting lesion localization," Expert Systems with Applications, vol. 36, pp.

2562-2566, 2009.

[21] V. A. Sitar-Taut, D. Zdrenghea, D. Pop, and D. A. Sitar-Taut, "Using machine learning algorithms in cardiovascular disease risk

evaluation," Journal of Applied Computer Science & Mathematics,

vol. 5, pp. 29-32, 2009. [22] S. U. Amin, K. Agarwal, and R. Beg, "Genetic neural network

based data mining in prediction of heart disease using risk

factors," presented at the IEEE Conference on Information & Communication Technologies, 2013.

[23] K. Srinivas, G. R. Rao, and A. Govardhan, "Analysis of coronary

heart disease and prediction of heart attack in coal mining regions using data mining techniques," presented at the 5th International

Conference on Computer Science and Education, 2010.

[24] E. AbuKhousa and P. Campbell, "Predictive data mining to support clinical decisions: An overview of heart disease prediction

systems," presented at the International Conference on

Innovations in Information Technology, 2012.

Dr. Salha Alzahrani has obtained her Doctor of Philosophy (Computer

Science) from the University of Technology Malaysia in 2012, and has conducted a research visit to the University of Oxford, UK in 2011. She

obtained her MSc with excellent pass (among the top 10 students) from

the University of Technology Malaysia in 2009, and BSc with honours (first rank) from Taif University in 2004. She is currently an assistant

professor at Taif University. Her contribution to the field of Plagiarism

Detection is illustrated by a number of indexed publications in high-impact journals and IEEE peer-reviewed conferences.

Afnan Althopity is working as a teaching assistant with the Department of Computer Science, Taif University. She obtained her BCs from Taif

University with excellent pass. She is teaching various lab courses and

introductory courses to computer science. She is also serving as a volunteer for the commnuty works such as Big Sister Little Sistser

conducted in her college.

Ashwag Alghamdi, Boushra Alshehri, and Suheer Aljuaid are

graduates from the Department of Computer Science, Taif University.



An Overview of Data Mining Techniques Applied for Heart ... · An Overview of Data Mining...

Documents

Transcript of An Overview of Data Mining Techniques Applied for Heart ... · An Overview of Data Mining...